mlsanityData to 14 Oct 2026

On the Robustness of Open-World Test-Time Training: Self-Training with Dynamic Prototype Expansion

arXiv 2308.09942 ↗First posted 19 Aug 2023cs.CV cs.LGcited by 34

Not on the map. Why?

Abstract (arXiv)

Generalizing deep learning models to unknown target domain distribution with low latency has motivated research into test-time training/adaptation (TTT/TTA). Existing approaches often focus on improving test-time training performance under well-curated target domain data. As figured out in this work, many state-of-the-art methods fail to maintain the performance when the target domain is contaminated with strong out-of-distribution (OOD) data, a.k.a. open-world test-time training (OWTTT). The failure is mainly due to the inability to distinguish strong OOD samples from regular weak OOD samples. To improve the robustness of OWTTT we first develop an adaptive strong OOD pruning which improves the efficacy of the self-training TTT method. We further propose a way to dynamically expand the prototypes to represent strong OOD samples for an improved weak/strong OOD data separation. Finally, we regularize self-training with distribution alignment and the combination yields the state-of-the-art performance on 5 OWTTT benchmarks. The code is available at https://github.com/Yushu-Li/OWTTT.

Key facts. Cited by 34 papers in the map (data through 14 Oct 2026). Builds on 33 of its 51 references that are in the map. First posted on arXiv 19 Aug 2023.

Builds on

33 of 51 references.

References · 33

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  3. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 7,005
  4. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,877
  5. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,223
  6. 2020FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidencecited by 2,085
  7. 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,917
  8. 2020The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalizationcited by 1,850
  9. 2020Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptationcited by 848
  10. 2018Deep Visual Domain Adaptation: A Surveycited by 635
  11. 2022Continual Test-Time Domain Adaptationcited by 525
  12. 2019Pseudo-Labeling and Confirmation Bias in Deep Semi-Supervised Learningcited by 510
21 more
  1. 2022Efficient Test-Time Model Adaptation without Forgettingcited by 469
  2. 2017VisDA: The Visual Domain Adaptation Challengecited by 435
  3. 2023Towards Stable Test-Time Adaptation in Dynamic Wild Worldcited by 415
  4. 2018Recent Advances in Open Set Recognition: A Surveycited by 310
  5. 2022Contrastive Test-Time Adaptationcited by 261
  6. 2019Adversarial Domain Adaptation with Domain Mixupcited by 250
  7. 2018Do CIFAR-10 Classifiers Generalize to CIFAR-10?cited by 200
  8. 2022NOTE: Robust Continual Test-time Adaptation Against Temporal Correlationcited by 151
  9. 2018Open Set Domain Adaptation by Backpropagationcited by 150
  10. 2021Cycle Self-Training for Domain Adaptationcited by 120
  11. 2021Source-free Domain Adaptation via Avatar Prototype Generation and Adaptationcited by 111
  12. 2019Infinite Mixture Prototypes for Few-Shot Learningcited by 99
  13. 2022Test-Time Adaptation via Conjugate Pseudo-labelscited by 98
  14. 2022Source-Free Domain Adaptation via Distribution Estimationcited by 97
  15. 2022Benchmarking Robustness of 3D Point Cloud Recognition Against Common Corruptionscited by 80
  16. 2019Discriminative Adversarial Domain Adaptationcited by 74
  17. 2022Reusing the Task-specific Classifier as a Discriminator: Discriminator-free Adversarial Domain Adaptationcited by 66
  18. 2022Evaluating the Adversarial Robustness of Adaptive Test-time Defensescited by 64
  19. 2019Open Set Domain Adaptation: Theoretical Bound and Algorithmcited by 48
  20. 2022Revisiting Realistic Test-Time Training: Sequential Inference and Adaptation by Anchored Clusteringcited by 44
  21. 2020Hypothesis Disparity Regularized Mutual Information Maximizationcited by 15

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 2; 2024: 11; 2025: 12; 2026: 9.

34 papers.

Sort:

By area:

  1. 2025Prompt-based Adaptation in Large-scale Vision Models: A Surveycited by 21
  2. 2024T4P: Test-Time Training of Trajectory Prediction via Masked Autoencoder and Actor-specific Token Memorycited by 20
  3. 2024Unified Entropy Optimization for Open-Set Test-Time Adaptationcited by 19
  4. 2024STAMP: Outlier-Aware Test-Time Adaptation with Stable Memory Replaycited by 13
  5. 2025Towards Robust Multimodal Open-set Test-time Adaptation via Adaptive Entropy-aware Optimizationcited by 12
  6. 2025Noisy Test-Time Adaptation in Vision-Language Modelscited by 10
  7. 2025Generalized Incremental Learning under Concept Drift across Evolving Data Streamscited by 7
  8. 2025Synchronizing Task Behavior: Aligning Multiple Tasks during Test-Time Trainingcited by 5
  9. 2023Continual Adversarial Defensecited by 4
  10. 2024Memory-Efficient Pseudo-Labeling for Online Source-Free Universal Domain Adaptation using a Gaussian Mixture Modelcited by 4
  11. 2024On the Adversarial Risk of Test Time Adaptation: An Investigation into Realistic Test-Time Data Poisoningcited by 4
  12. 2025Towards Understanding Extrapolation: a Causal Lenscited by 4
22 more
  1. 2024Efficient Open Set Single Image Test Time Adaptation of Vision Language Modelscited by 3
  2. 2025Test-Time Adaptation with Binary Feedbackcited by 3
  3. 2023Mixture-of-Experts for Open Set Domain Adaptation: A Dual-Space Detection Approachcited by 2
  4. 2024COMET: Contrastive Mean Teacher for Online Source-Free Universal Domain Adaptationcited by 2
  5. 2024Exploring Human-in-the-Loop Test-Time Adaptation by Synergizing Active Learning and Model Selectioncited by 2
  6. 2024Covariance-based Space Regularization for Few-shot Class Incremental Learningcited by 2
  7. 2026TTL: Test-time Textual Learning for OOD Detection with Pretrained Vision-Language Modelscited by 2
  8. 2026Dual Distribution Estimation for Zero-shot Noisy Test-Time Adaptation with VLMscited by 2
  9. 2024Improving Batch Normalization with TTA for Robust Object Detection in Self-Drivingcited by 1
  10. 2025From Pixel to Mask: A Survey of Out-of-Distribution Segmentationcited by 1
  11. 2025Instance-Aware Test-Time Segmentation for Continual Domain Shiftscited by 1
  12. 2026TALON: Test-time Adaptive Learning for On-the-Fly Category Discoverycited by 1
  13. 2026Mitigating the ID-OOD Tradeoff in Open-Set Test-Time Adaptationcited by 1
  14. 2024Open-World Test-Time Training: Self-Training with Contrast Learningcited by 0
  15. 2025Should Bias be Eliminated? A General Framework to Use Bias for OOD Generalizationcited by 0
  16. 2025Hierarchical Adaptive networks with Task vectors for Test-Time Adaptationcited by 0
  17. 2025Segmentation Assisted Incremental Test Time Adaptation in an Open Worldcited by 0
  18. 2026PlugSI: Plug-and-Play Test-Time Graph Adaptation for Spatial Interpolationcited by 0
  19. 2026DynProto: Dynamic Prototype Evolution for Out-of-Distribution Detectioncited by 0
  20. 2026COSTA: A Cluster-Centric Paradigm for Annotation-Free Open-Set Semantic Segmentation of Aerial Point Clouds with Domain Shiftscited by 0
  21. 2026Reliable Neural Collapse Approximation for Open-World Test-Time Adaptationcited by 0
  22. 2026Test-Time Generalized Category Discoverycited by 0